This is the code for "WToE: Learning When to Explore in Multi-Agent Reinforcement Learning". The paper will be published soon.

1. Grid Examples: The environment contains two basic grid environment (2-room and 4-room environments), which are implemented in the GRID/ENV file.
2. Multi-Agent Particle Environments (MPE): A simple multi-agent particle world with a continuous observation and discrete action space, along with some basic simulated physics. Used in the paper Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments. We test our algorithm in six scenarios, including simple_adversary,simple_crypto, simple_push, simple_reference, simple_spread, simple_tag.
3. MAgent: MAgent is a research platform for many-agent reinforcement learning. Unlike previous research platforms that focus on reinforcement learning research with a single agent or only few agents, MAgent aims at supporting reinforcement learning research that scales up from hundreds to millions of agents.

Please follow the instruction of 'README.md' file in different environments to install Python requirements.

@article{WToE,
  title={WToE: Learning When to Explore in Multi-Agent Reinforcement Learning},
  author={Dong, Shaokang and Mao, Hangyu and Yang, Shangdong and Zhu, Shengyu and Li, Wenbin and Hao, Jianye and Gao, Yang},
  journal={IEEE Transactions on Cybernetics},
  year={2023}
}